Which test for a rate vs heavy-tailed revenue?
Test selectionMediumProblem. For the same checkout experiment you have two success metrics: payment-success (a rate) and revenue per account (ARM, which is heavy-tailed — a few members spend far more). A teammate runs a Student's t-test on both. What test would you use for each, and why is the t-test on revenue risky?
Before you reveal: say your answer out loud, as if you were in the real interview — get your reasoning across clearly first. There is no single correct answer: reading what the interviewer is really after and defending your own thinking is what makes an answer strong.